Smartphone application for detecting and visualizing basal stem rot disease stages in oil palm

Introduction As one of the world's most important oil crops, oil palm plays a crucial role in meeting global food and industrial demands. However, Basal Stem Rot (BSR) disease poses a severe threat to oil palm plantations, significantly reducing yields and shortening plantation lifespans particularly in Southeast Asia. Effective early detection and monitoring are crucial for mitigating its impact.Methods This study presents an integrated approach to BSR disease stage detection and visualization through a combination of smartphone applications, deep learning-based classification, and a Web GIS-based dashboard. A dataset comprising images of oil palm trees in healthy, early infected, and severely infected stages was collected using a dedicated smartphone app. Five state-of-the-art convolutional neural network (CNN) architectures: DenseNet201, InceptionV3, MobileNetV2, NASNetMobile, and ResNet50 were evaluated for classification performance.Results MobileNetV2 emerged as the best-performing architecture, achieving an overall accuracy of 77% while balancing accuracy with computational efficiency. This model was subsequently integrated into the smartphone application "Gano Stage" for real-time disease stage prediction. The app enables plantation managers and relevant stakeholders to monitor disease progression, with predictions automatically updated on a Web GIS-based dashboard for spatial analysis and decision-making.Discussion The proposed system demonstrates practical utility, scalability, and adaptability, particularly in resource-constrained environments. By offering an accessible and efficient early detection solution, it contributes to the sustainability of oil palm plantations.